Travis Katz is the CEO of BrightDrop, a subsidiary of GM that makes electrified delivery vans with an eye toward rebooting everything about how delivery works. BrightDrop has pretty big partnerships already, with names like FedEx, Verizon, and Walmart committed to its Zevo 600 van, and it’s got big ideas for making the steps from the van to your door more efficient as well with something called e-carts.
Katz says there’s a huge demand for delivery, especially as online shopping keeps getting bigger, but the transportation network is at capacity, and you can’t just keep throwing more trucks and drivers on the road or making city streets wider. His plan is to redesign the entire system to make it more efficient. So I wanted to know how he’s attacking that problem and making it manageable, all while getting buy-in from customers who won’t really accept delays or increased costs.
BrightDrop is a wholly owned subsidiary of General Motors, so I also wanted to know how that works, what he gets from being part of the big company, and which parts slow him down. Lots of classic Decoder stuff in this one.
Okay, Travis Katz, CEO of BrightDrop. Here we go.
This transcript has been lightly edited for clarity.
Travis Katz is the CEO of BrightDrop. Welcome to Decoder.
Hey, Nilay. Thanks for having me. It’s great to be here.
Yeah, I have a lot of questions for you. BrightDrop does a lot of really interesting work in the delivery world. Let’s start at the beginning. What is BrightDrop?
At BrightDrop, we are reimagining how goods arrive at your front door in a way that’s smarter, more sustainable, and safer. When you talk about delivery specifically, a lot of people are like, “Okay, that sounds boring.” Let me tell you why it’s not boring. This is something that touches every one of us. Think about the Amazon packages you have delivered to your door daily; every single person in this country is ordering stuff online daily. It’s a $5 trillion market today, so it’s huge.
It’s growing really fast, but we’re also in a time where the wheels are kind of starting to fall off in the industry. It has grown so quickly over the last few years that we’re starting to run into real challenges around how we continue to scale it. The infrastructure of cities is fixed, and there’s a strong desire from everyone to get growing carbon emissions under control. We’re really setting out to help solve a lot of these problems with the products we’re building.
We have a suite of products in three broad categories. We have electric vehicles — specifically, we are building electric delivery vans. We have a big one, the Zevo 600, and a smaller one, the Zevo 400. We have a series of electrically propelled containers or carts. These e-carts are called the Trace. We have one specifically for package delivery and another one for grocery delivery. Then we have a software suite called BrightDrop Core that ties it all together.
We’re working with some of the biggest companies in delivery, like FedEx, DHL, Verizon, and Walmart. We’re really thinking about how we reimagine this space for an all-electric future.
With that suite of solutions, you have the two vehicles — the vans and the carts, which are really interesting — and you have the software that ties them all together. You just gave a pitch that’s like a startup pitch. It was really well-rehearsed. “Here’s the size of my total addressable market. Here are the products we can make. Here are the customers we can get.” You aren’t a startup, right? You’re a wholly owned subsidiary of GM. How did this all come about?
I’m a startup guy by background. I’ve been working in the internet space. I was one of the early guys at Myspace, I helped launch the News Corp digital group, and I also started my own company, Trip.com, which I led through a successful acquisition back in 2017.
GM has an innovation incubator, like a lot of large companies do. There’s a group of people thinking, “We have all this great technology and all these great assets, so what are some new businesses we can create?” Typically, when big companies have these incubators, they’ve come up with lots of ideas that rarely get traction for lots of reasons, which I’m happy to talk about. I think that was historically true for a lot of things that GM worked on as well. When they started this one, they said, “Look, we need to take a new approach. We need to bring in somebody from the outside, who has actually built a business from the ground up, to lead this.”
I got the call from GM a couple years ago to pitch this idea. Then, when I was two weeks into the role, I looked around and said, “Look, we have these great ambitions, but there’s no way I can build the business you’re wanting to build, set up the way that it is, that runs the way General Motors runs.” So we set it up to run as a separate company, as a standalone startup. We are a subsidiary of GM, but we run fairly independently from GM. We set it up in this way to get the best of both worlds. I can leverage all of the assets that General Motors has, which are many in terms of vehicle design, manufacturing, supply chain, and battery technology, but I can also run at the speed of a tech startup.
So far, that has been a pretty killer combination.
I promise you I want to talk about BrightDrop and the business, but everything you just said is pure Decoder bait. It’s a structure conversation. How does a big company structure itself to be more nimble and get the value out of the scaled businesses? I’m constantly saying Decoder is secretly a podcast about org charts. Here’s the most complicated org chart of all, and I’ve heard so many companies say they’re going to do this. “We’re going to have an internal startup that can run faster than the bureaucracy of the big company.” You just mentioned that a lot of times that doesn’t work. I’m curious why you think it doesn’t work, and what you’ve done in this structure to make sure it does.
The reality is that the needs of the startup are very different from the needs of the large company. If you think about a company like General Motors in particular, General Motors has been around for 114 years, and that entire time they have been doing one thing really well, which is building cars. Building cars is fantastically complicated. They have a system in place that allows them to launch 30 new models in a year. They can build these cars in factories all over the world, managing supply chains. It is a tremendously well-run company. It’s run through a set of rules, procedures, policies, and decision-making forums that are specifically designed for that.
“You can’t run a software business on a car trajectory.”
There is a lot of goodness going on there, but if you’re building software, you build it very differently than the way you build a car. The risk profile is different and the speed at which you iterate is quite different. You can’t run a software business on a car trajectory.
You also get the risk tolerance. If you think about a large, publicly traded company, particularly one that’s in a mature market, you’re really looking at how you’re going to meet quarterly earnings. It’s very focused on profit, points of margin, those sorts of things. When you’re starting up a company, you need to be thinking a little bit further in the future. How am I going to grow this business to be a large, high-margin business in the future and invest now to get there?
The fortunate combination we had here was that this was not General Motors’ first rodeo with this. They own a company called Cruise, which is the autonomous vehicle subsidiary. I think there was a lot of learning through the Cruise process of how this works.
When we came to BrightDrop, we borrowed a lot of the structures that were used for Cruise. We had to go through and think specifically about the areas that we needed to run very differently and what areas we didn’t. We had to have a lot of discussions about why we needed to learn to work differently. Thankfully, General Motors has a really visionary CEO in Mary Barra. She learns very quickly and she got it right away. She was like, “Just tell me what you need to do and let’s make it happen.”
So you’re a startup person. Mary hires you to be the CEO. What do you say you need in that conversation? Everyone knows about the graveyard of big company startup incubators in the past, and it is quite a graveyard. What did you ask for to make it work, and what did she give you?
A lot of things. One, I needed to have control over a lot more decision-making. One of the ways that General Motors works is that it has a lot of centralized functions and centralized decisions. Purchasing, supply chain, manufacturing, engineering, and design are all centralized. The instinct is to say, “Well, we have all these resources there and they’re good people who know what they’re doing, so let’s leverage all of these central resources to build the business.”
The kinds of products we’re building look different from cars, and so I needed to bring in talent that could build the kind of business we wanted to build. So we needed to think about talent very differently and about what compensation and structures would attract the right kind. We needed to be able to take decision-making out of this central structure and decentralize it into BrightDrop so that we could very quickly make decisions, iterate, and move at rapid speed — and we’ve been doing it.
The speed at which we’ve been building this business has been pretty spectacular. We first announced this business in January of 2021 at CES. We delivered our first vehicle that year, and we delivered it in 20 months from conception to delivery. That’s the fastest General Motors has ever created a vehicle. I think it may be the fastest any company has ever brought a vehicle to market. We’ve been hitting records since then. We got a world record for the longest distance driven by an electric delivery van. Our vans get about 250 miles, so the range is quite good.
I think Andy Hawkins from The Verge was in that van when you set the record actually.
Yeah. Andy Hawkins was there for the whole thing, which was really fun.
We announced in November that we are on target to be the fastest company to hit $1 billion in revenue. We’re aiming to do that in three years from launching the company. To put that in context, it took Tesla 10 years to hit $1 billion. I feel like this combination, this startup backed by a large company, is paying off pretty well for us.
I have a lot of questions about where that revenue comes from and how sustaining it is, but I want to just stay in the structure for one more minute or two. You’re a startup person. Usually the reason people are startup people is because they start a company, take on a lot of risk, and then they have an exit, make a lot of money, and then do it all over again. I know that for a lot of people who come on this show, that is their history. Is there an exit inside of GM? You said you had to change who you recruit and how. Are you recruiting people against the possibility of an exit event like that, that would make them rich?
Not in the traditional sense. We’ve had to structure compensation in a way that we can attract the kind of people who work in tech companies, without having the promise of an exit. At the same time, we have set up the company as a standalone subsidiary. If we made the decision that we wanted to raise external capital or do something else, that’s something we could do in the structure we have. But right now we’re working with GM, and they are a company that generates a lot of free cash flow, so we don’t need external capital. We have to structure in a way that attracts the right kind of talent.
Probably job number one when I came in was to really sit down with the leadership team at GM and talk about how tech companies think about talent. What talents do we need that don’t exist in the company today? What do we need to do to bring them in? That wasn’t a quick conversation. We had to go through a few rounds, but like I said, Mary is a pretty sharp woman, so we got there pretty quickly. We’re attracting great talent from Uber, Lyft, Google, and all the tech companies you would hope you’d be able to attract.
You said you might be able to raise external capital. Is there any thought that GM might spin BrightDrop off into another public company?
We don’t have any plans to do that today. If times change and the company decides they want to, it’s something that’s possible. But right now, it’s just heads-down building the business.
Let’s talk about BrightDrop itself and its structure for a second. How is BrightDrop structured? You don’t need a bunch of battery engineers, right? GM has that platform.
Probably the simplest way to think about it is that we are running like a tech company. We have a hardware business, a software business, and a vehicle business.
For the vehicle side of the business, the best analogy is that GM sort of works like a contract manufacturer for BrightDrop. We say, “Look, this is what we need from the vehicle and what we need to be able to do,” and GM does that. For everything else, we’re building software that runs in the vehicle, desktop software, mobile software, and e-carts for different verticals. All of that is happening at BrightDrop, so we do actually need battery engineering and that sort of stuff for the e-carts.
I guess the simplest way to put it is that we’re structured very much like a typical tech company. We have product management, design, and engineering for both hardware and software. We have a whole sales organization where we go to market and engage with customers, to help not just sell the products, but to also really think about how we can partner with them to solve problems. It’s very unique.
In some ways this is an area that really sets us apart. There are a lot of other companies that are bringing to market electric vehicles for the delivery space. The reason they’re doing that is because it’s a massive space. E-commerce is a $5 trillion market, and it’s going to be growing to $7 trillion by 2025. There’s a big opportunity there, so there are a lot of people saying, “Hey, we’ll build a van to help these companies deliver stuff.” We’re not thinking about a van as a standalone product. We’re thinking about an ecosystem of products. We’re really thinking holistically about the challenges that these delivery companies are facing.
Let me explain what those things are. As you can imagine, there have been companies delivering stuff to our door for a long time, but the pace of that has been growing dramatically. With the rise of Amazon and the rise of e-commerce, we’re seeing just massive double-digit growth year after year after year.
At the same time, we’re seeing expectations from consumers of, “I want things to arrive faster and faster.” First, it was like, “Okay, if I can get it in a week or 10 days, that’s great,” and then Amazon came out with next-day delivery. Now there’s a whole push for, “Can I get it delivered to my door in an hour or two?” That’s really, really challenging. It requires rethinking the entire system. How do you stage goods closer to people? How do I organize and track where goods are and predict what goods are going to be needed in certain areas? There’s a lot of complexity there.
In addition to that, to keep pace with this growth, companies have been saying, “Okay, let’s just throw more bodies and more trucks at the problem.” The challenge we’re running into now is that the infrastructure of our cities is fixed. You can’t make the streets wider, so as you put more and more delivery vans into urban areas, you start to see massive challenges with congestion. We’ve all seen this. These delivery vans can’t find a parking space, and a lot of times they’re double-parked and blocking a lane of traffic. This happens at multiple points in a city simultaneously, which is creating logjams and making people late to work or late to pick their kids up after school.
It’s also dangerous. A lot of times, couriers are having to step out into traffic to get out of these vans and bicyclists are trying to get around the van that’s sitting in the bike lane. The current system doesn’t scale. We’ve reached the maximum scale.
“How can we reimagine how these goods get there?”
What we’re trying to do with BrightDrop is really say, “Let’s take a step back and look at this problem holistically. How can we reimagine how these goods get there?” Things like the Trace e-cart that we built have been tested with FedEx in multiple cities, and we’re testing it with a bunch of other partners as well. FedEx was able to deliver 25 percent more packages per day using the Trace than without.
Let me describe the Trace for your listeners. Imagine essentially a big box or a locker on wheels with electric propulsion. It can carry 250 pounds of packages. With the electric propulsion, it feels effortless. Instead of dragging along this dolly with things strapped onto it, it’s effortless, it’s easy to maneuver, and it fits in an elevator. When a truck pulls up, instead of doing five trips back and forth to the van to serve a single high-rise building, you can do the entire thing in one go.
You can also start to reimagine how you do it. Do you create a central drop-off point or a micro-hub where you can have trucks drop these things off and give them to bike couriers or foot couriers and fan those things out across the city? To do that, of course, you need a lot of software. Ultimately, delivery is like a massive optimization problem, where you have tons of packages that need to get to tons of people in tons of different places, and you need to figure out what’s the most efficient route. A lot of the software we’re designing is helping to reimagine how you do this at the lowest cost, with the lowest possible carbon emissions, and at the highest speed.
It’s a really fun problem. It’s complicated, but we are the only company out there that’s really taking this holistic look at the problem. Our customers really appreciate it when we have these conversations. They’re like, “Oh great, you actually do more than just sell electric vans. You really understand our business.”
The carts are super fascinating to me. Every time I look at the pictures on the website or read about what you’ve done with them, it strikes me that you’re containerizing the last mile of delivery, the way that container ships revolutionized ocean freight. You’re making the problem more modular and more understandable for software. Right now, all the boxes are different shapes and sizes, and it seems like these carts converge them into a standard that you can then have software optimized for. Is that right?
It’s exactly right. If you’ve studied the history of containerization and shipping containers, it’s really fascinating. In the old days of shipping, a ship would pull into port and it would take days to unload everything and stuff would get spoiled or lost. It was really complicated.
Then they came up with the concept of a shipping container. Everything was pre-packed. You could crane these things off of a ship onto the back of a truck or a train cart. You could package, modularize, and move everything very fast. So a lot of what we’re doing with the carts is exactly what you’re describing. It’s taking this concept of containerization and bringing it to the last mile, but with a sort of modern spin on it.
We’re in the era of connectivity, internet of things, so you can start to have much better tracking of where things are. We can have auto lock and unlock to improve security, so you don’t have things like package theft, which is a huge deal in last-mile delivery. You’ll be delivering things in malls and people will grab packages off of the dollies and run away with them. We increased security and we increased the chain of custody visibilities, so you can see who opened this and when. We can track where things were as they move throughout the chain.
It also unlocks this way of reimagining how we plan our routes. What we’ve seen from a lot of the modeling that we’re doing is that you can start to use things like e-bikes, e-carts, and other forms of transportation to allow you to deliver the same amount of packages with fewer vehicles.
A lot of people say, “Well, why would you want to sell fewer vehicles? Isn’t that one of your revenue streams?” The reason we want to do it is we want to make the system work right. If we can make the system work more efficiently and better, everybody wins. It’s a massive opportunity and it’s one that’s very timely, because like I said, the system is starting to show cracks. We just can’t keep throwing trucks at the problem.
The thing that strikes me, every time I talk to anyone on the show who has sort of a hardware / software split, is that the investment in software just rapidly begins to outpace the investment in hardware. My favorite example of this is John Deere, the tractor manufacturer. Their CTO came on the show and they’re like, “Yep, we spend more money on software than hardware,” which is just unbelievable to me. Are you on that same trajectory, where the software side of the business will become a much bigger investment and require more resources over time than the actual EV vans?
I think we definitely have more software engineers than we have hardware engineers. The difference with hardware versus software is that there are a lot of capital investments with hardware. If you’re going to build a hardware product, you need a factory with tons of robots and working capital to buy all the parts. It’s very expensive to launch a hardware product versus a software product, which you can actually launch pretty cheaply.
“A really great software engineer is not 50 percent more productive ... they’re five or ten times more productive than an average software engineer.”
You do need good engineers. One of the things very unique to the software world is that it’s not linear from a mediocre software engineer to a great software engineer, it’s an exponential curve. A really great software engineer is not 50 percent more productive than an average software engineer; they’re five or ten times more productive than an average software engineer.
I’m a big believer in trying to hire really, really good people and fewer people, as opposed to just hiring lots of bodies. Ultimately, the reason everyone is putting all this investment in software is because software is really changing the game, changing the way that hardware products work, and changing the capabilities of what they have.
This is happening in the auto industry right now. I think Tesla was the first one who really understood the potential of what you could do with software in a vehicle. Everybody else is investing heavily to get to the bar that Tesla set. It’s exciting. In the old days, cars were essentially these mechanically engineered products; they had engines, you put gasoline in them, and they went. There’s some software for things like safety systems and automatic braking, but they kind of ran in a disconnected way. When you start connecting them, it just unlocks tons of value. You can make the car better through software over time as opposed to your car losing value quickly year by year. You can add new features and functionalities, you can optimize charging…
Wait, can I ask you a really dumb threshold question?
I have a lot of auto CEOs on the show. I’ve heard that set of statements many times. If you want to tell Mary to come on the show, she’s welcome to make the same set of statements.
One thing that I’ve always wondered about is something I’m curious about with BrightDrop as well. There’s a connection that everyone makes between, “We’re going to go to EVs,” and, “The car is a computer. We’ll get all the value of computerizing and connecting the car.” There’s actually not, as far as I can tell, an articulated connection between those ideas. Like, why does the drivetrain of the car matter to whether or not you can add more software features to the infotainment? I suppose that’s not relevant for BrightDrop, but do you understand what I’m saying? Everyone says the two things are going to happen at once. Even for BrightDrop, it’s like you’re going to do EV vans and you’re going to do optimized logistics management and containerized packaging so that the battery-powered trolleys can work. There’s not necessarily a connection between those two things.
No, you’re absolutely right. You can build smart connected cars with internal combustion engines. I think part of it is just sort of a coincidence of timing that we’re living in the era of software in the moment when the economics of electric vehicles suddenly became viable. So part of it is just a timing issue.
That said, there’s a lot you can do with batteries. When you think about value creation for a consumer, you save a lot of money when you move from an internal combustion engine to an EV. If you compare a BrightDrop Zevo van versus a traditional diesel van, if you fully utilize it and you’re driving it to the full charge, you can save on average $7,000, $10,000 per year on fuel cost alone, because electricity is so much less expensive than diesel. That’s the starting point. There’s a lot of power that you can get because these things are connected to managing the charging. I’m going to get into utility geek mode here for a minute.
Let’s do it. This is what the show’s all about. Let’s go for it.
So, anyone who lives in a state like Texas or California, which is where I live, you have occasional rolling blackouts, and those are happening because there are too many things drawing electricity from the grid at the same time. If it’s really cold in Texas, everyone is turning on their heat and suddenly there’s not enough power going onto the grid. If it’s a really hot day, everyone is turning on their air conditioner. Our electricity grid in the US is old and outdated, and we as a country need to invest in making it stronger. In the interim, utilities are doing things like paying you money if you shut off your electricity use for 10 minutes when they’re having a demand surge — meaning there’s more demand than there is supply on the grid.
For a delivery vehicle, most of these guys are out delivering packages during the day and then they come back to the depot to charge overnight. They’re sitting there for eight to 12 hours, but they only need a couple of hours to be fully charged. If you have the ability to turn off your charging for 10 minutes when there’s a surge because everyone’s turning on their air conditioners, that’s actually a revenue opportunity for the vehicle producer and for the owner of those vehicles.
That’s just an example of the kinds of things you can do with connected vehicles. You can also send electricity back into the grid and power your home. There really are a lot of cool things when you marry software with electrification that didn’t exist before and don’t really exist with an internal combustion engine. You talk about things like, “Hey, can I make my infotainment system more compelling and add features?” You could do that with an EV or an ICE vehicle. If you look at how GM is building their software stack, they are doing the same kinds of software across both types of vehicles.
You have a unique perspective. You’re close to a big company, but you’re sort of outside of it. Are people just excited about batteries, so they’re willing to reconsider everything? Do you shove everything into that opportunity, or is there a rational connection? It seems like with the logistics — and charging in particular — there’s a rational connection, but then everything else just comes along for the ride.
When I look at why we are building BrightDrop now and not five or ten years ago, it’s really about the convergence of three big trends. The first is the maturity of EVs. For years and years, people have been talking about EVs, but the cost of the batteries was so high that it just didn’t make economic sense to own one. You could buy it because you wanted to produce less emissions, but it was going to cost you more to drive that. You were sort of donating your cash to help the environment. We have crossed the tipping point, where it’s now cheaper to drive an EV than an internal combustion engine, which is why EVs are suddenly hitting primetime.
At the same time, with connectivity and the internet of things, we suddenly have a fast wireless network. You can connect to the internet to control cars, cell phones, sprinkler systems, and so on. That is unlocking all sorts of new businesses and new technologies.
Then the third one is big data and machine learning. When you start to have all these connected devices, they start to stream lots of data back into the cloud, and machine learning has now hit its sweet spot. You’re seeing this with ChatGPT, where you can use software and algorithms to create a lot of value, generate insights, and optimize what you’re doing in a way that you never could before. A lot of us take it for granted, but we’re living in a very exciting, unique time in the history of the world where these massive changes of technology are all converging and unlocking massive potential. I think one of the ultimate applications of that is autonomous driving.
GM owns this company called Cruise, run by a guy named Kyle Vogt, who is absolutely brilliant. Those three things — electric vehicles, big data, and connectivity — can actually create cars that can drive themselves. They’re no longer prototypes, and they are doing commercial rides with no driver in them. In San Francisco, you can order a taxi with no driver today. It’s real. It’s a super exciting moment to be living in. The automakers are uniquely positioned to take advantage of it because they know how to build cars at scale, and that is a very hard and capital-intensive thing to do. Very few people have the skillset to do that. You’re seeing this kind of renaissance and revolution, and it’s also a race to say, “Okay, who’s going to become the leaders in this next generation?”
Is the plan for the BrightDrop vehicles to be driverless — to drive themselves, stop in the middle of the city, and let people go deliver packages?
We haven’t announced anything specifically in terms of our autonomous plans, but I think we are big believers in autonomy as a whole. We think there is a big role to play for autonomous vehicles in the delivery space.
There are simple dock-to-dock applications that are fairly easy to tackle. When you start to think about delivering things to people’s front doors, there’s a whole host of complications about how you solve the last 100-foot problems. If I’m on the 18th floor of my high-rise apartment on a Zoom call with my boss and my new toothbrush arrives, what do I do there? When we think about these smart motorized containers, we’re thinking about, how do you marry those with autonomous vehicles to unlock the power of autonomy and the opportunity for savings autonomy brings for the mass market? We’re doing a lot of thinking there, although I don’t have anything specific I can share today.
Whenever I talk about EVs to a car executive who’s focused on consumers, the first question my audience asks is, “Charging sucks. How are you going to make the charging better?” Obviously, a fleet approach, an enterprise approach, doesn’t have the same set of challenges, but you still have to install a lot of charging infrastructure. That’s a big startup cost. How are you handling that?
Charging infrastructure is a big challenge. People ask me this a lot, and the way I put it is that in the early days of the internet, when people talked about concepts like streaming video or even streaming audio, everyone was like, “Yeah, that’s cool, but the network can’t handle streaming audio and streaming video,” but of course we got there. I think the technology to do charging at scale in America is there. It’s going to take a little bit of time and it’s going to take investment — in the same way as it took investment in our cellular networks to allow you to do streaming video from your phone — but it’s coming. It’s not going to be a long-term major barrier, but it is a short-term pain point.
For commercial fleets, it’s a little bit different than consumer vehicles. For these fleets, the typical pattern is vehicles are out in the field for eight hours delivering packages and they’re back charging in the depot overnight. There’s no need for in-the-field charging of delivery vehicles.
Well, you’re still going to install 40 chargers at a depot. That’s an expensive investment that requires a lot of power in one place.
Correct. Not to minimize, there are still some major challenges there. You have to install the chargers. A lot of times if you’re going to charge 50 vehicles at once, you need more power coming into that depot than you may have today, which might mean you need the utility to come and upgrade their network there. A lot of times the depots where you’re parking these things are not owned by the delivery company, but they’re leased, so there’s a landlord involved. There is a lot of complexity here the companies are working through.
When we go out to sell these vehicles, we try to talk about these challenges early because you actually have to plan ahead for thinking about where you want to deploy these vehicles and how you get the charging infrastructure in place. We have a whole bunch of partners. GM put together partnerships with a lot of the major charging infrastructure providers, so we can bring those guys in at a moment’s notice. GM has also launched its own energy business, so it’s building its own charging infrastructure. It has storage and there’s a whole fuel cell component where you can do stationary storage with fuel cells to do load shifting and that sort of stuff.
We have a whole bunch of solutions that we can help make this easier for customers. But the reality is you have to plan for it. You have to look at your schedule and how you want to roll these things out nationally. Most of the companies, as they’ve all made commitments to go to EVs, are starting to spin up divisions and teams to look at this.
Like I said, it’s going to be a bumpy road for the next couple of years as everyone figures out and gets smart about how this works, as the utilities make sure they have enough staff to handle all of the interconnects, and that sort of thing. We’re sort of charging into uncharted territory here and there’s going to be a lot of learning along the way, but none of the problems are that technically hard, it’s just mostly a question of planning and resources.
We’ve done 100-plus episodes of the show, you’re the first person to come on and say 5G is important to you, even in a roundabout way. So that’s exciting. Everyone else is like, “Yeah, that’s this thing that happened.”
I want to talk about the car maker part of it. You have vans, the Zevo 600 or the 400, and they’re big. You refer to General Motors, your contract manufacturer. We’ve seen around the industry that there are two approaches to making an EV. There’s ground-up, which is what you, Rivian, Lucid, and Tesla have done. Then there’s the F-150 Lightning. “We took our truck platform, we took all the value of the scale from that platform, and we added batteries to it. So we’re not going to literally reinvent seats, we’re not going to reinvent instrument clusters, we’re just going to swap out the drivetrain, take advantage of the scale, and sell people a product that they really like.”
Why engineer your vans from the ground up? GM has vans. You can go buy a GMC Sierra van or whatever it’s called. Why not take their existing commercial vans and electrify those? Why go from the ground up?
Yeah, that is an excellent question. You either stuff some batteries into your existing gas-powered vehicle and say, “Okay, it’s an electric vehicle,” or you can say, “Let’s start from scratch, from first principles, and say we’re going to build an electric vehicle. How do you build the right kind of vehicle leveraging that platform?” That’s the approach that we took at BrightDrop, and that’s the approach that GM is generally taking with all of its products. Not to be disparaging to anyone else, but that is the right way to do it.
When you build something from the ground up as an EV, it allows you to really design the vehicle around this form factor. There are a lot of benefits to doing it. If you’re sitting in the backseat of a traditional car, there’s that lump in the middle of where your feet are supposed to go because you have an axle there. You don’t need that in an EV, so you can get rid of things like that.
You can also really start reimagining things. For example, because we have this battery that sits under the floor of the vehicle, it allows us to do things like lower the step-in height of the van. We took the step-in height down by several inches versus a traditional van. If we had an internal combustion engine, you need to have axles and all that stuff under there. People say, “Well why does that matter?” Well, the average courier delivering packages is stepping in and out of those vans 100 to 150 times per day. There’s wear and tear on your knees when you’re getting in and out. If you can lower that step-in height by a couple inches, it just makes it a much better job. It’s a lot less strain, a lot less injury, a lot less tiring, and you don’t have churn of drivers.
You can also maximize the space that you’re using for the battery to get better range. What we’ve seen from a lot of automakers who try to just stick batteries into their gas-powered vehicles is that those vehicles don’t perform well in the field. They get substandard range and can’t really meet the requirements that our customers need in the field. They’re really struggling to sell a lot of them. We’ve heard a lot of this — that the performance just isn’t there. Our strategy of saying, “Let’s start from the ground up. Let’s build around this electric platform,” allows us to both build a better solution for customers, but also one that performs better and will hold its value much better over time.
So the vans are designed from the ground up, they make for a better product for the drivers, and then you sell them to delivery companies along with a pretty advanced software suite called BrightDrop Core. I’m curious, when you go to FedEx, which is a delivery company, they obviously have a massive software operation of their own. They obviously care about logistics a lot. You say, “Hey, we’re going to swap out the trucks. Not only the trucks, we’re going to try to standardize the actual units that you’re shipping to fit into these carts. Then on top of that, we’re going to sell you software to make that all work better.” How do you make that sale? That’s their business, and you’re starting to go directly into what they do. How do you make that sale and make it sound like you’re not just commodifying what they do?
Our goal is not to commodify it.
You have FedEx and DHL. They’re pretty intense competitors, right?
Right. The reality is that they’re good at what they do. The reason those are the biggest delivery companies in the world is that they know how to do this well. They’ve figured out how to scale this business and be able to deliver these massive millions of packages per day to customers in a way that’s cost-effective. The starting point for them is, “We don’t need your help.” Yeah, that’s true. They want electric vehicles, so all of these companies — whether it’s FedEx, Walmart, or Amazon — have committed to getting to net zero from a carbon perspective by 2035 or 2040. They need help with electrification. The starting point is, “Hey, we have vans that are designed around the way that you do business.”
To make this clear, there are traditional cargo vans that you see driving around and then there are delivery vans. Delivery vans look very different. They have these high seating positions so it’s easy for couriers to get in and out of the vehicle, and they have a big bulkhead door so they can go from the driving cabin to the back to get packages. Our designers spent hundreds of hours in the field doing runs with real couriers to look at how they are working during the day and how they are managing packages. So we’ve built a product that’s perfectly designed for the use case that these guys have, but it’s electric. It’s a pretty compelling value proposition going in with the vans.
The next step from there, with things like the e-carts and with the software, is that we have to be able to prove that these things create value in the field and that they’re easy enough to adopt that they can move them into their operations. A lot of what we do when we go and talk to these customers is talk about the problems we know they’re facing, how our solutions can help them solve those problems, and then we pilot them. We get out in the field and we try it.
So like I said, with FedEx, we have been testing these in multiple markets. We’re measuring the results. We start off measuring the baseline. How fast can you deliver these packages per day? How many stops per hour? These kinds of things. Then we test again when we put our solutions in there. Do you get better results? Most times we see that, yes, they do. That’s sort of the starting point, but it’s not a quick sale.
To build a business like this, to build any enterprise business, it is a multi-year partnership. You have to really partner with these companies and help them through the process, because it’s not just about buying new pens or printers or things like that. You need to rethink your operations to incorporate these things into the business. Every day they wake up they’re like, “Oh my God, we have to get all these packages out the door to all of these people before 10AM.” They don’t have lots of free time where they’re sitting around just twiddling their thumbs, looking to try new things. You have to figure out how to make it easy to adopt and start slowly rolling it out so they can develop a playbook. Once they have it working in one depot, then can they try it across three depots? Can we then scale it to 10? It’s a multi-year journey we’re on with a lot of these customers.
On the software side, a lot of it comes from taking the software and connectivity that we have in these things at the starting point and saying, “Hey look, here’s what the data is telling us about how you’re working, and here are opportunities for you to do more with less, get more done, take more advantage of these electric vehicles, think about your charging differently, and think about how you could incorporate these e-carts to allow you to scale to more packages without adding more vehicles at all.” It’s a really interesting conversation. We have to start by showing the value, and then from there we can move into talks about how to adopt them.
When you talk to a FedEx, it seems like the real key here is the e-carts. It’s the standardization so they can use different kinds of delivery vehicles once they get the trucks out there. You keep talking about, “We can’t just keep adding more trucks.” How do you go to a FedEx and say, “All right, in order to make the e-carts work, we need to standardize and containerize the actual shipping units”? That is an uncrackable problem from my vantage point.
It’s not straightforward, honestly. Like I said, these guys have a way of working, in the same way that GM has a way of working. They’ve designed it and perfected it over many, many years, and that way of working doesn’t include using e-carts. But they are quite aware of the pain points.
I have this photo, and it’s representative of something that happens in multiple points over New York City, Chicago, Miami, LA, and everywhere else outside an Amazon-owned Whole Foods. There is a truck dropping off packages and taking up not only an entire lane of traffic, but also the curb, the lane where people are supposed to be parking, and the sidewalk. There are packages spread out everywhere. A traffic jam is created as a result as people and bikes try to get around. It’s an incredible mess that they’re creating, because they have so many packages to move and there’s no place to park these vehicles.
Even though they have a system that works for them and they’re getting by, they know the system isn’t perfect and has challenges. If we can bring them solutions to the real pain points that they’re feeling in the field, they’re all ears. They’re happy to hear it and they’re happy to try things.
I think the nice thing about this business is that we are singularly focused on this problem of how we get people’s packages to their front doors. That focus allows us to really understand the pain points and design solutions around them. For them to adopt them, they have to change some of the ways that they’re loading things into the vehicles. They have to think about the order they’re putting packages into these e-carts, and that’s a software problem. It’s not something they can do everywhere all at once. They need to start small and build out from there. That’s exactly what we’re doing.
You said you’re on track to $1 billion in revenue. Is that coming from sales of the trucks? Is it coming from sales of e-carts? Is that by selling software services and that growing faster over time? How’s that broken down?
It’s coming from all three. We will get to $1 billion from revenue selling all three of those products in 2023. In the early years, we’re going to see more revenue from the vans. Then over time, the balance will shift increasingly towards the e-carts and the software.
The van is the more mature product. We started on it earlier, so it’s a little bit more of a known product, and it’s faster and easier for a customer to adopt. They’re switching out a van powered by gasoline to a van powered by electricity — but it’s a van. They understand what a van is. The software and the e-carts are new products, and we’re still bringing those to market, but we’re a little behind the van in terms of manufacturing that stuff because we started it more recently. That business is going to ramp up pretty quickly. That is what we expect to see and all indications are a go right now.
Do you think that you’ll eventually subsidize the price of the van against the expected margins of the software business? Right now, TVs are really cheap. They’re cheaper than they should be because they run a lot of connected TV services. If you buy a Roku TV, there’s an advertising unit built into it that comes along for the ride. The actual price of hardware is really quite subsidized by a service stack. Are you foreseeing the same sort of thing happening, where you sell the van at a discount or even under cost because you’re going to make it back up on the software?
“One of my favorite things about software businesses is that they sort of scale infinitely.”
I’m not sure. I think there’s definitely potential. One of my favorite things about software businesses is that they sort of scale infinitely. You have a set of fixed costs, which is people, servers, and storage. You can create one set of software and then scale that out over a lot of customers without very much variable cost. Variable costs are near zero, so you tend to see very, very high margins in software. Yeah, those margins over time can become a competitive weapon that you use to help drive more sales of vehicles and other hardware. So I wouldn’t say “No, we won’t do that,” but right now we’re not really thinking about it that way.
Do you attach an expected services revenue to the sale of every van, or are you just saying, “Okay, vans are a mature business; people understand them, they want to buy them, and they have committed to electrification by 2030. We need to get a product to meet the demand, and then hopefully we can sell a software solution on the back of it with infinite scalability and potentially infinite margins”?
One of the goals we set at the beginning of the year is that every hardware product we sell is going to be bundled with a software solution. That’s the way we’re building the business. As a customer, you want the software because the software creates value for you. It’s not like we’re trying to force software on you. We’re going to use software to help you save money. The sell is pretty straightforward and pretty easy. I don’t expect we’ll be selling hardware products that don’t come bundled with software.
When you think about a FedEx and a DHL, the big customers, they want to be differentiated. What are the points where you can tailor the product to their needs?
Generally, when you’re selling software to an enterprise as opposed to consumer software, you’re going to have to do some level of customization by default, because you have to integrate into all of their systems. So FedEx, DHL, whoever, all have their methods of running their business. They have their own software stack, which is usually some combination of stuff they’ve built themselves and third-party tools that they’ve integrated in. When you go to them, you have to really figure out, “Okay, how do we integrate our software with yours to make it work in the way you want to run your business?”
Our goal is not to say, “Hey, here’s how you run your business.” Our goal is to say, “We’re going to help you run your business better. We will empower you with the data and insights to help you do things smarter, but we’re not going to tell you how to run your business.” We don’t know their business as well as they do, obviously. So it’s really more of a supporting function. We’re not necessarily building custom tools for customers. We need to figure out how to make it plug in, which inevitably requires some level of custom work.
The elephant in the room is Amazon
It’s hard to talk about potential customers for BrightDrop without talking about the elephant in the room, which is Amazon. I think the simplest way of describing their retail operation or store is as a software-based logistics platform. They had an exclusive arrangement with Rivian, which also has an electric delivery van, but that has come to an end. Is that an opportunity for you? Did you spin up a strike team to go take the business from Rivian at Amazon?
We are talking to everyone, and from day one we’ve been talking to every player in the business who is looking to move goods. If you look at how most of these fleets buy today, the reality is that they rarely go single-source and say, “We’re only going to buy from a single manufacturer.” They tend to spread their bets around and pick a couple of key partners that they want to work with. I have yet to see a fleet that operates with only one partner. So yeah, we’re talking to everyone. I expect we will be selling vans to everybody that moves goods out in the market.
Amazon is the one customer I can see saying, “Look, we don’t want any of your software, we just need the vans. We’re really good at software.” Would that be the customer that says, “Okay, we’re just going to sell the vans. That’s a big enough opportunity, and we can try to do something else everywhere else”?
Yeah. Like you said, Amazon is great at software — that’s their core — but at the very least, they need connectivity and data to be able to run the software that they’re designing. Amazon is still thinking it through. They have lots of money and lots of talented people, so in theory, they could build anything. I think they need to decide what the core stuff is.
I don’t think they want to get into, “Okay, we’re going to own and maintain our own vehicles and try to figure out the maintenance schedule for electric batteries.” They’re going to still need software from us. They’re welcome to do that if they want, but I don’t think it’s where they need to be spending their time. Their time is really figuring out their own logistics system and how they move most efficiently and at the highest speed as part of their competitive advantage.
I think there are lots of opportunities for us to work together with them, but we may work with them in different ways than other companies. How much of the software stack they are going to want may look different from other customers. I think it may look a little bit different for every customer. But there’s a core functionality there that they’re really going to want and need. The economies of scale of having one player building it and being able to sell it to multiple vendors is advantageous, versus building it all just for your own operations.
Do you look at Rivian’s challenges in scaling and say, “Whew, it’s a good thing that we have GM behind us,” or do you face the same challenges?
Well, if I look at the shift that’s happened in the EV startup world over the last 12 months, it’s been pretty dramatic. We went from a world in which capital was almost free, so there was money sort of everywhere and it allowed a lot of companies to go after a space like automotive, which is incredibly expensive to bring to market. It’s incredibly hard and takes a long time. Then suddenly, as interest rates started going up, the entire sentiment changed and valuations of startups have been crushed. It’s hard.
When I look at where BrightDrop is at, we don’t share a lot of the struggles the startups are having in the space. A big challenge everyone is dealing with right now is how to scale. The question of supply chain has been a big challenge for the automotive industry.
GM has massive know-how here. We know how to scale vehicle production, and we are scaling it as we speak. We have our factory up in Ontario, Canada, and we are rapidly ramping that up and rolling vehicles off the market. We don’t have a lot of the challenges. That really comes down to GM’s backing. They know how to manage batteries, how to build EVs, how to scale production, and how to manage supply chain and supply chain risk. It’s really hard trying to build all that stuff from scratch.
Elon has talked a lot about how hard it was to get Tesla off the ground and how they almost went bankrupt multiple times. It is a very hard space to do it and you need a lot of capital. There are some great startups out there and some really smart people and I would like to see a lot of them do well, but it’s going to be a tough 18 to 24 months in that space. I think we’re going to see a lot of companies really challenged.
Even GM is challenged though, right?
GM has made a lot of promises about how fast they will electrify and what the cars will be, but there’s only two or three of them right now. The Hummer EV barely counts. It’s a concept car that actually shipped, which is remarkable in its own way, but it’s not meant for everybody and it’s not shipping at scaled volumes. When you think about, “All right, there’s a crunch on battery supply and GM has huge plans for consumers,” where does the negotiation for BrightDrop come in? “All right, we have $1 billion in revenue and we’re growing. We’re going to take priority on GM’s battery supply”? How does that work?
General Motors is launching a lot of EVs over the next couple of years. It’s an interesting moment in history because you’re seeing this giant rush where all of the major automakers, as well as the startups, are trying to launch lots of EVs at once, and the supply chain is not totally ready for that. Everyone is trying to figure out, “Okay, how do we scale as quickly as possible?”
I think General Motors is better-positioned than most other traditional auto manufacturers because they started much earlier and with a model of vertical integration. GM has its own battery platform, the Ultium battery platform, which is a joint venture with LG Chem. They started investing in building their own battery plants in the US — domestic supply, which matters for incentives and other things. They have also been focusing on, “Let’s build vehicles from the ground up.” They just started much earlier than everyone else, which puts them in a much better position than most of the competition.
Is the BrightDrop van on the Ultium architecture, and are you going to get the benefit from that?
Yes, we are using the Ultium batteries. It is true that to scale all of these things, it is hard even for a company like General Motors to scale very quickly on a totally new platform. There are still kinks being worked out. I have an inside view to see that the company is doing very well, so I’m very confident, but there does have to be a prioritization of where the batteries go.
I think General Motors looks at BrightDrop as one of the key growth drivers for the company, so we are getting our fair share of batteries. I expect we will continue to do so over time. Even the big companies are having to do startup things and go from zero to one when it comes to scaling EVs. There will be bumps along the road for sure, but I think the long-term path looks quite bright.
If you have more demand than the Ultium division can supply, are you allowed, as a division of GM, to go source other batteries?
Are we allowed? Yes. Would we want to? Probably not.
The Ultium platform is a really great platform, and you’ve probably seen some stuff about it. It’s an overall architecture that allows you to configure different battery sizes and shapes, and you can use different types of cells: pouch cells, prismatic cells, or traditional cylindrical cells. It’s a very flexible architecture.
With the way vehicles come together, you have to have the battery software and the battery architecture integrated in with all the other components. It’s not the same as if you have your remote control for your TV that you can just swap batteries in and out and it doesn’t really matter what the brand is. They’re pretty tightly coupled with the overall architecture of the vehicle, so it’s not simple to switch them out. Could we do it? Yes, but I don’t think we have any reason to. The Ultium platform is going to be scaling pretty fast and it gives us what we need. I don’t think there are better batteries out there.
The last two questions are the kind of big-think questions. You started off early in the conversation by saying you really wanted to do something in climate change. Obviously BrightDrop, EV company, delivery, there’s a lot to do with climate change there. I’m looking at a stat here that says, “The current pace of electrification in the transport business is not fast enough to prevent the worst effects of climate change.” Do you feel pressure on that side of your brain to go even faster and to push even harder?
The thing that gets me motivated every day when I wake up is that this business has to succeed. Not just because we want it to succeed, but because the planet needs it to succeed.
We absolutely are in a state of crisis from a planetary perspective. We’re seeing the effects of climate change with tornadoes, the flooding in California, the droughts in Colorado, or the fires that are happening. These effects of climate change are affecting everyone, and these disasters are getting stronger. We really need a concerted effort. We’re working on one piece of the pie and there’s a lot of other people working on a lot of other pieces, but we need it to succeed and we need it to move very quickly.
Inside the company, the people that are working at BrightDrop are people who want to take their smarts and their skills and their passions and apply it towards solving a problem that matters, as opposed to saying something like, “Okay, how can I optimize the click-through rate on an ad in my Instagram feed?” It’s a very motivating force for me. It’s a very motivating force, I think, for our team.
There’s a lot of work that still needs to be done, but I’m very optimistic. We actually have the technologies in place now to make a big dent in climate change. You’re seeing technology sprouting up across the ecosystem to deal with other parts, not just transport, but also electricity storage or electricity production. I’m very hopeful that we have the technology in place, we just need focus, and our business is very focused on that.
General Motors as a company is very focused on driving this change to net zero. Mary Barra was the first CEO of a major manufacturer who came out and said, “We’re going to go all-electric for consumer vehicles by 2035,” and has put the capital behind it to make it happen. There’s a lot of work still to be done, but we have a lot of smart people and a lot of passion behind it, and I think we are going to have a major impact. Will it be enough to stop the worst effects? I’m not sure, but I think we can get there.
You’ve sketched out a really big vision for how you want BrightDrop drop to work. You containerize the packages that go on the trucks, those trucks park in the middle of the city, then an army of worker bees on e-bikes and motorized carts takes the packages to their places, come back to the trucks, and then the trucks leave. We’ve made everything cleaner and more efficient. Is that happening yet? Have you run that test? “Okay, we want the biggest vision we have to work all the way end-to-end.” Have you seen it happen?
Yes. The one we can talk publicly about is the test we did with FedEx, where they were able to deliver 25 percent more packages per day using these Trace e-carts than without. Normally when you bring a new technology to an enterprise, getting 5 percent or 6 percent better is a huge win. You never see numbers like 25 percent. In fact, they were so big the first time we looked that we said, “We must have made a mistake. Let’s run the numbers again — and then run them again.” The numbers kept coming back, so we ran this test in other cities and saw similar things.
It really is game-changing, and it’s just the starting point. That really was not leveraging all the power that the software has. We acquired a company last year called Marain, which came out of Stanford’s Autonomous Systems Lab. It built a whole bunch of technology around optimization of routes and charging, and really applying machine learning and simulation to help. You can essentially run 20,000 combinations of your city’s routes in an hour or so and come back with better solutions.
As we’re starting to run this with partners, we’re seeing major opportunities to do things better. It’s pretty eye-opening when you have a company that’s been running a business for a long time and thinks, “Hey, we have this pretty well-nailed,” but then you can come back to them with data that says, “Hey, there’s actually a smarter way to do this, and you can get more value.” You can show them what that data looks like and how you can make those changes. It’s pretty exciting. Like I said, we’re at the very beginning of this journey, so there’s a lot more to come, but I’m very optimistic where we’re sitting right now.
Let me bring that down to the present. That’s the vision. You ran the test, and you saw the numbers and where it could go, and there’s stuff beyond there. Right now it’s about going to customers and making it happen. What’s next for BrightDrop? What’s the next concrete step that helps you get to that vision?
We spent the last two years building prototypes and testing things out with customers in the field, and 2023 is the year where we’re moving from science projects, laboratories, and experiments to actually commercializing these products. We are going out to market and engaging with all of the biggest players in the space, with the vans, the e-carts, and the software. Like we said, we’re aiming to do $1 billion of revenue this year. That’s a big mindset shift for the team, and everyone is very excited and motivated by the opportunity.
If we take a look forward, we’ll be launching our second van. The Zevo 600 is live now, and we’re going to be launching the Zevo 400 later this year. When you think about the technologies that are out there, like our sister company Cruise, what happens when you marry autonomous driving with delivery? That’s pretty interesting. When you think about containerization and how you marry that with robotics, you start to think about a driverless vehicle deploying containers or lockers where you can pick up packages on the sidewalk.
There’s so much exciting stuff happening in technology, and it’s all super applicable to this space of getting all the things we order online to our doors in a way that’s fast, safe, smart, convenient, and has zero emissions. The sky’s the limit.
We introduced a new product last fall for grocery, which we’ve been testing in the field with Kroger. I think grocery is another very interesting space because it’s been exploding since the pandemic in terms of the number of orders. There’s tons of inefficiency; most grocers are losing money from grocery delivery. There are smarter ways to do it and meet that customer demand. We’re excited about that space as well. So, there’s lots more to come.
That’s great. Well, Travis, thank you so much for being on Decoder today. It was really fun talking to you.
Thanks, Nilay. It was a great pleasure to be here. I really enjoyed the conversation.
Decoder with Nilay Patel /
A podcast about big ideas and other problems.